SQL, analytics and data engineering / Benchmark leaderboard
SpreadsheetBench / Verified
Real spreadsheet manipulation; full release has 912 questions and Verified has 400 instances.
No published company result yet.
This benchmark is in the research directory. Its task package, adapter and grading protocol need qualification before a hosted run can be offered.
Arrange an evaluation for your company →What this benchmark measures
Metrics
Output workbook/cell correctness under the published evaluator.
Execution requirements
Agent capable of editing spreadsheet files.
Scope and limitations
A SQL-only product is not directly compatible. Declare full versus Verified and workbook tooling.
Evaluation availability
Catalog entry. Request a managed evaluation to qualify your agent interface and the benchmark’s native grading requirements.
Sources and company fit
- Original benchmark ↗Established specialist
- Dataset / catalog ↗
Companies whose products may fit
Research recommendations based on product capabilities. These companies have not necessarily run this benchmark or integrated with Blobfish.
Compatibility notes for each company
Anthropic: Capability-aligned; adapter/access to qualify
Google DeepMind: Capability-aligned; adapter/access to qualify
Hebbia: Capability-aligned; adapter/access to qualify
Julius AI: Capability-aligned; adapter/access to qualify
Microsoft Copilot: Product-specific adapter / qualified access
OpenAI: Capability-aligned; adapter/access to qualify
Rogo: Capability-aligned; adapter/access to qualify
Shortcut: Capability-aligned; adapter/access to qualify
Hex: Conditional on a configured workflow accepting the required workbook and producing the native graded spreadsheet output.